Over the past ten years, increasing attention has been given to how workers express emotions in a variety of work settings (Ashforth and Humphrey, 1995; Rafaeli and Sutton, 1987, 1989; Sutton, 1991; Wharton and Erickson, 1993). An underresearched, yet critical, aspect of the literature on emotions in organizational life concerns employers' attempts to control and direct how employees display emotions to customers. Emotional labor, generally defined as the act of expressing organizationally-desired emotions during service transactions (Ashforth and Humphrey, 1993; Hochschild, 1983), is the central focus of this study. This article seeks to extend previous theoretical and empirical research on emotional labor in four ways.
First, a more rigorous conceptualization of emotional labor is presented. By drawing on previous emotional labor studies, psychological and anthropological research on emotions, and impression management studies, a three-component conceptualization of emotional labor will be advanced. The framework presented here suggests that emotional labor can best be described in terms of frequency of emotional labor, duration of emotional labor, and emotional dissonance experienced as a result of having to express emotions one may not actually feel.
The second objective is to identify the organizational and job characteristics which might predict emotional labor. Previous researchers (Adelmann, 1989; Ashforth and Humphrey, 1993; Hochschild, 1983; Wharton, 1993) have suggested, but rarely tested, variables which may help to predict which work roles will require regulation of emotional expression and what conditions might influence employees' willingness and ability to express sanctioned emotions.
The third objective is to explore the consequences of performing emotional labor on employees' well-being. Previous research has implicitly or explicitly concluded that emotional labor has negative and dysfunctional consequences for workers (Adelmann, 1989; Erickson, 1991; Hochschild, 1983). This study suggests the possibility that under certain conditions, performing emotional labor actually leads to favorable attitudinal and role behavior outcomes.
Finally, the article examines the implications of this research for more effective management of emotions during service transactions. The rapid and significant increase in the number of jobs which require regulated displays of emotion, as well as the potential impact of emotional displays on service quality and customer satisfaction, certainly makes this issue one worthy of additional attention.
Conceptualization of Emotional Labor
According to Hochschild (1983), jobs involving regulated displays of emotion possess three characteristics: (1) they entail voice or facial contact with the public; (2) they require the worker to produce an emotional state or reaction in the customer; and (3) they provide the employer an opportunity to control the emotional activities of the employee. Displaying organizationally-sanctioned emotions to customers or clients has been argued to be a form of "labor" since it requires effort, planning, anticipation, and adjustment to situational factors in order to publicly display emotions that employees may not necessarily privately feel (James, 1989).
Frequency of Interaction. A categorization of jobs requiring emotional labor provided by Hochschild (1983) established the foundation from which virtually every existing empirical study of emotional labor has since proceeded. The premise here is that external stakeholders (customers or clients) are more likely to comply with organizational goals when the affective bonds of liking, trust, and respect have been established through appropriate employee behavior. Thus, the more a work role requires contact with other people, the greater the organization's need to rely upon regulated displays of emotion to ensure compliance with organizational goals. …